Printable View
Didn't even know there was a forum for Cube3 modders :)
I posted the picture of the E3D hotend on Facebook hoping to produce some attention for the Cube3 printer. The proprietary nature of the device casts it in a bad light and I felt like offering a full conversion to an open source machine would make it at least a little more popular. Can't really beat what you get for ~$100 these days.
I'll be releasing the E3D hotend files, along with numerous other files, very soon. The last work I have to do (other than a bit more testing) is relating to endstop placement. I have the concepts worked out, just need to turn them into reality now :)
I can't wait to see this in action!! It would be really cool to be able to use off the shelf parts to extend the life of this printer. I actually get really good prints with it now and I worry what I will do when it dies lol. Parts are going to be scarce for it soon I'd imagine.
Quote:
Originally Posted by Cubistic
Here's a link to the carriage design. More to come in time :D
https://www.thingiverse.com/thing:2762341
now just look at what you've done. Going to make me convert mine! This look fabulous. Cant wait to see what the final results bring. What type of tweaking would need to be done to the computer part if any?
Technically it can be used with the default but you'd need to setup a proper resistor to provide the computer the feedback it's looking for. I'm going to be using a RAMPS board with itQuote:
Originally Posted by ztoddman1
Goodness gracious, this is exactly what I am working on right now! Except, my goal is to have 2 extruders and keep the Cube 3 as nice looking as it came out of the box, including the touch screen. Initial goal was to keep the electronics, but finally I gave up on this idea as the board of the Cube 3 is plain stupid. So, for now the goal is to make it work with 32bit MKS SBASE 1.3 (running Smoothieware) plus MKS TFT24. Apparently TFT 24 has exactly the same screen that Cube 3 has, which makes things much easier. It must have the heated bed too. The parts are already ordered and I'm hoping to have them by the end of the next week.
Here are my designs:
Attachment 2501Attachment 2500Attachment 2503Attachment 2502Attachment 2499
Some initial prints and assembly:
Attachment 2505Attachment 2507Attachment 2506Attachment 2504
Playing around with MKS SBASE board placement :) That is probably the best way to place it, but will require grinding off some parts of the metal frame.
Attachment 2508
Currently I'm fighting my OCD regarding the placement of the optical endstops :D Probably placing them on double-sided scotch tape is not a great idea :D Comments and suggestions are welcome!
Ditch the optical endstops and use TCM2130 instead. Sensorless endstops via stepper motor feedback
Interesting idea! But the board already has DRV8825 drivers installed, in this case I'll have to use the external drivers. My concern is already limited space plus this approach may not be very good for the belt teeth health. That's the way Cube 3 original board was dealing with the endstops, but it was not doing it very well, the sound of skipping belt teeth drove me crazy :DQuote:
Originally Posted by Cubistic
@Kiza. Is that you Alex?
G70
Yep, thats me ^_^Quote:
Originally Posted by JackMeoff
How about flags under the base for the "Y" motion end stops? Should remain within the confines. The floor of the "tray" seems to have the most amount of room.Quote:
Originally Posted by Kiza
This is the endstop I'm going to be using:Quote:
Originally Posted by TommyDee
Attachment 2510
So far thinking of using the longer M3 socket cap screws that are holding the rails on the frame to attach the plastic holders for the endstops, still have to design them, waiting for the actual endstops to arrive.
The tray is going to be used to hold the board itself, so no room there. I will have to be creative with the placement of the board.
- - - - - - - - - -
Some heatbed testing in action :D
Attachment 2513Attachment 2514
Looks like you need to print one of these...
https://www.thingiverse.com/thing:2644528
:)
Case is great! But I will not use this controller in the final version, since I'm going to use the MKS SBASE 1.3 board, which has a heated bed control. This controller I bought when I wanted to reuse the Cube 3 original board. But not anymore.Quote:
Originally Posted by TommyDee
- - - - - - - - - -
I am still struggling with a good and clean placement of the optical endstops, specially on the Y axis. Really don't want to over-complicate it.
The main problem is that a lot of good places are covered with the shell parts.
You can move the sensor with the y-tray and fix the flag below. Loosely carry the wires with the heater wires.
Thinking of this too, just cannot figure out yet the good placement and the way of placement. Really don't want to cut or drill any part of the shell itself.Quote:
Originally Posted by TommyDee
@Kiza
You crack me up!!!
Do you always do electronic testing with your cloths off? Is that a safety thing?
G70
Hahahah, made my day ^_^Quote:
Originally Posted by JackMeoff
- - - - - - - - - -
Parts are finally here!
Some pictures:
Attachment 2534
This is just plain G O R G I O U S !!!
And a video, no endstops yet:
https://youtu.be/Cr4zULB28Ow
If you would go the TRinamic route, you could avoid having to add endstops.....
But AMAZING JOB! Congratz. The sound of the raising bed is even alien! haha
Thanks a lot! ^_^ Well, sensorless would be great, but TMC2130 is not yet supported by smoothieware, as far as I know...Quote:
Originally Posted by bolsoncerrado
Could be my future upgrade!
Before you convince yourself you need sensors... have you confirmed that it is actually a belt-jump happening for homing the Cube 3?
I suspect it just hits the torque limits of the motor and what you are hearing is the motor, not a jumping belt.
Just a thought.
Could be, I just feel safer with sensors. A friend of mine is printing the endstop mounting brackets for me, so hopefully today I'll do the first proper homing! ^_^Quote:
Originally Posted by TommyDee
You are having way to much fun, Kiza! :cool:
In fact, yes, I am ^_^Quote:
Originally Posted by TommyDee
Endstops:
Attachment 2545Attachment 2546
Z-endstop still needs some work though. Does not fit when transparent extruder cover is installed.
Homing:
https://www.youtube.com/watch?v=8d7GBzVWg8k
- - - - - - - - - -
By the way, does anybody know, what are the settings for movement/extrusion? Steps per mm in Cube 3 firmware.
The bearing is 10mm OD... so (Pi*10)/(step angle*360) minus the allowance for filament crush in the driver "gear" since it represents a slightly smaller OD. This could be significant. So you could base the diameter on as little 9.5mm OD.
New Z endstop, corrected. Everything fits perfectly now!
Attachment 2562
What, no colorful heat-shrink on the solder terminals? :p
Maybe turn the y-axis switch 90*CCW (tray switch)? With wires, that may just be cleaner.
From an ME point of view, we like to let the switch signal but not interfere.
However, you are in a tight spot on both X and Z.
And you have hard stops.
- - - - - - - - - -
oh, and maybe connect x and z to a single plate so you don't need to disturb the rail screw.
Of course in the final version they will be there! ^_^Quote:
Originally Posted by TommyDee
Moved it in already, otherwise the nozzle was stopping right on the edge of the table. That provided more space for the cables.Quote:
Originally Posted by TommyDee
That's right, have no idea how to solve it. Tested for impact, they hold just fine. Thinking of enabling the limit switch option in Smoothieware, but that triggers general error, needs restarting every time.Quote:
Originally Posted by TommyDee
That was an initial idea, but then I felt so lazy to do that :D Rail screw is just fine, I have placed a longer one.Quote:
Originally Posted by TommyDee
The z-hard stop is the nozzle, sadly. I would consider an adjustable mechanical stop just below the nozzle crash.
A z-sensor really doesn't make sense past the hard stop so you need to match the hard stop with the actual trigger from the switch.
You could use the switch for a slightly different function, such as a "slow-ahead" signal that can have steps programmed in after that point.
You might have a bounce problem in the switch if you try to set a difference between crash and the 0.07mm gap.
Then again, maybe it is not a home signal. Maybe it is a motor cut off when everything else goes awry, including the hard stop.
What is the DIY crowd's interpretation of how to properly use the Z-signal?
R u gonna do an Instructables when you're finished?! :D
Well, whatever I have seen so far is accomplished the same way I have, that's the most common DIY approach.Quote:
Originally Posted by TommyDee
I have tried to crash the Z-caret to Z-endstop, nothing fatal happened so far.
Adding more sensors requires the support from the software side, which is currently not in place. I am still doing more research on this.
Of course smoothieware is open-source, so we can write whatever we want, to add the functionality you mentioned above. But that's gonna be the later stage for me.
Currently I am researching the "Soft endstops" feature, which limits the movement after the homing procedure is done. Looks promising, but didn't work so far for me.
- - - - - - - - - -
Of course, once I am satisfied with the results! ^_^ Not much left, I need to finally assemble the hotends and tune them. Haven't done it before.Quote:
Originally Posted by bolsoncerrado
Plus, the challenging part would be aligning and leveling the nozzles. Once it is done, I will run some tests on the open frame, and move on with assembling the printer.
Looking forward to what you come up with. And I am not advocating an additional sensor, just playing with the logic of the available gap.
Finally solved the problem with soft endstops. My bad, I was using the old firmware.
Now everything works! No more grinding sounds ^_^
But now my concern is, how to use the wiper blades to clean the nozzles.
Video:
https://youtu.be/YTBOeHnA6AY
Formula 1 speed ^_^:
https://youtu.be/bG25cCB7tng
You don't need the wipers anymore. That is a 3DS solution for selling more consumables.
Remember that ring every other machine draws before printing the part?... That's the "purge".
- - - - - - - - - -
So where is the Z-switch triggered?
Sorry, forgot to capture this moment. Will do next time.Quote:
Originally Posted by TommyDee
I actually liked the idea of placing the printhead above the bin to purge some filament, before starting the print :) But I can live without it :)Quote:
Originally Posted by TommyDee
No worries.
I'm still trying to zero in on the "normal" DIY Z-homing works from the hardware perspective.
So here is me convincing myself Z-home would be when the platen is at the bottom?
That would make the upper switch an "emergency stop function" if you implemented it.
This makes sense to me if you think about lead screw setups.
Cube 3 has the advantage that the bottom is the bottom (home) when you un-power (and un-short) the stepper motor.
The number of steps between the bottom (home) and a proper filament gap to the plate is the variable that counts in Z.
- - - - - - - - - -
Can't you write that into a gcode format and use that as the part's gcode prefix (setup routine)?Quote:
Originally Posted by Kiza
I'd say that would be something a lot of gcode printer people would be fond of.
Yeah, but in this case I either change the horizontal size of the print table, and my 0 point is outside the real table, or, I need to find out if I can enable/disable the soft endstops using gcode.Quote:
Originally Posted by TommyDee